A popular metric is to compare CEO pay to the pay of the lowest-skilled–and hence lowest-paid–worker in the company. If the CEO makes $15,000,000 and a burger-flipper makes $15,000 then the CEO is making 1,000 times more.
This provides fuel for policy debate. As I often say, socialism is not about economics. There is no economics of socialism. It is nothing more than institutionalized envy. And the pay ratio of 1,000 appeals only to envy. It is offered without context, as none is deemed necessary. The very fact that he is paid so much more is taken as proof of … well it’s outrageous.
Something must be done!
Or so the Left argues. Get him because he’s rich. Impose a pay restriction on all companies, or at least all public companies. Tax the #$&%! out of him. Give all workers a so called guaranteed basic income (so called because socialism never delivers on its promises).
Libertarians are correct to note that if a wage is set in a free market, there is no fairer mechanism. And besides, it’s the company’s money to spend as it chooses. No good can come out of Washington meddling. They also point out that the CEO produces more value than the burger flipper, and his job is harder and more stressful.
This is true, but the very notion of a ratio of one person’s pay to another’s is meaningless. It tells us exactly nothing, while tantalizing us that it tells us something important.
I propose a different metric. A ratio of a worker’s pay to the value he produces. For example, suppose the gross profits (ex. cost of meat and ingredients) generated by the burger flipper’s work is $15 per hour. His pay is $7.50. Therefore this is a 2:1 ratio. He is paid half the value he creates.
Now consider his manager. The manager supervises 10 workers. Assuming all have equal productivity, the total gross value of these workers is $150. If the manager is paid $15, then this same ratio moves up quite a lot. It is 10:1.
Finally, let’s go all the way up to the CEO. If the firm’s gross profit is $15,000,000,000, then this ratio is 1000:1.
As we move up in the value chain, it is harder to measure the gross value contributed by each worker. For example, with the CEO, is the gross value simply gross profit? Or are we trying to look at just the difference he makes compared to his predecessor?
However, it should be clear that as you go up in value and salary, the ratio of value created to salary also goes up. That is, higher-value workers are paid less as a percentage of the value they create. Isn’t that a different view from the common envy view?
I am not sure what to call it. It may even exist, though I have not come across it. I am mulling “gross value capture ratio of wages”.
Keith Weiner Economics 
Great piece. Probably would be hard to calculate, as you say. I’ve seen some analysis a long time ago that accounted for size of enterprise. Of course a CEO of a larger enterprise should be paid more. Perhaps that could be ratio, CEO compensation compared to market cap.
BTW, an investor I have little patience for griping about CEO pay in the abstract, but I demote companies which engage in shenanigans re CEO pay. For example, if the CEO has too much power in regard to compensation committee. Does compensation shift rules after the fact when it comes to calculating incentive pay? Just because lefties target CEOs with envy does not mean that we libertarian types should ignore agency risk in which management shafts shareholders. I’ve got no problem with a CEO who makes a trillion dollars, so long as the performance matches the comp.
Oh yes, agency risk is real. But it’s a problem for the owners, not for politicians!
In the early 60s Elliott Jaques https://en.wikipedia.org/wiki/Elliott_Jaques proposed the concept of time-span of discretion, which simplistically is how long until the results of an person’s work/decision is known (delay until feedback is received). If a buger flipper undercooks the meat or flips it to the floor, they and those monitoring the flipper will know within a few minutes. For a CEO making a strategic decision to enter into a new line of business – success may not be know for years.
Jacques found that what was perceived as equitable payment by all people across an organization was closely related to the time-span of discretion, ie people had an intuitive feel that CEOs should be paid more because of the complexity and inherent uncertainty of their jobs and decisions made.
Needless to say, I can’t see that his research ever got developed, as CEOs aren’t interest in funding someone to see if his research was applicable across many companies and could come up with a scientific rule on relative pay rates.
Saying the agency risk is for owners not politicians ignores the fact that politicians are the ones empowered to vote on laws that govern companies. Do owners really have a https://en.wikipedia.org/wiki/Say_on_pay ? when board directors are usually ex-CEOs and director pay is related to CEO pay.
The jobs with the longest time-span are the ones that require the most judgement and have the greatest potential to create value.
I don’t think politicians should have that power.
I certainly agree that politician shouldn’t have that power, but currently the laws those politicans have set don’t give that power to owners either. It is a problem for the owners, but the problem is political, not just one of the owners controling CEO remuneration.
How do the owners not have the power to set or reset CEO pay?
As per the wikipedia link, the “shareholders [have] a non-binding vote on executive compensation” – non-binding. Shareholders only have “agents” in directors who are supposed to control the “agent” executives, two levels of agency where both groups come from the same “class” and who’s interests are aligned regarding increasing pay. The ability to vote directors out is difficult for a diverse group of shareholders to do, particularly when you have crony capitalists like Buffett and institutional investors controling large share blocks, and is only an indirect method of remuneration control.